A modified cyanide-nitroprusside method for quantifying urinary cystine concentration that corrects for creatinine interference

Clinica Chimica Acta 289 (1999) 57–68 www.elsevier.com / locate / clinchim

A modified cyanide-nitroprusside method for quantifying urinary cystine concentration that corrects for creatinine interference Y. Nakagawa*, F.L. Coe Department of Biochemistry & Molecular Biology, and Kidney Stone Program, Department of Medicine, University of Chicago, Chicago, IL 60637, USA Received 12 February 1999; received in revised form 19 July 1999; accepted 14 August 1999

Abstract Cystinuria, an inherited disease, is clinically diagnosed by detecting cystine in urine. A colorimetric method using sodium cyanide and sodium nitroprusside is a simple qualitative test used to detect cystinuria. Several colorimetric methods have been proposed for quantitative analysis of cystine; however, we found that none of them were satisfactory because the results were not reproducible. The causes of non-reproducible results were: (1) insufficient reduction time for conversion of cystine to cysteine, and (2) the interference of creatinine. In this report, we present a method to quantitate cystine in urine. We also found that ascorbic acid and ferric chloride, but not zinc chloride, interfered with the color reaction. Using this method, 15 normal urine samples (10 males and 5 females) and 12 cystine stone forming patients’ (5 males and 7 females) urine were analyzed. The method was compared to commercially available urine controls. Only captopril showed a dose dependent response and color intensity at 521 nm. Thiola and D-penicillamine showed little effect on cystine determination.  1999 Elsevier Science B.V. All rights reserved. Keywords: Cystine; Cystinuria; Colorimetry; Nephrolithiasis

*Corresponding author. Fax: 1 1-773-702-5818. E-mail address: [email protected] (Y. Nakagawa) 0009-8981 / 99 / $ – see front matter  1999 Elsevier Science B.V. All rights reserved. PII: S0009-8981( 99 )00159-X

58

Y. Nakagawa, F.L. Coe / Clinica Chimica Acta 289 (1999) 57 – 68

1. Introduction Cystinuria is an inherited disease in which reabsorption of cystine and three basic amino acids (arginine, lysine, and ornithine) in defective [1]. Cystinuria patients are genetically classified into three types (Type I, II, and III). This classification depends on a complete or incomplete recessive phenotype. In addition, allelic mutations, either homozygotes or heterozygotes, result in variable phenotypes [2–5]. This disease is most often the result of a genetic mutation in the gene designated as SLC3A1 located on chromosome 2 (2p21) [2]. Its incidence has been reported to be between 1:15 000 and 1:7000 [6]. Gitomer and Pak reviewed published articles on inherited cystinuria, and described 21 mutations and nine polymorphisms in the SLC3A1 gene [7]. More recently, Rozen et al. screened 23 cystinuria children in a Quebec screening program, and characterized their disease type, kidney stone formation, and specific mutation of the SLC3A1 gene [8,9]. Since the solubility of cystine is low at normal urine pH (5.0–6.5), the patients may develop cystine stones. It is clinically important to analyze cystine in urine qualitatively and quantitatively. To detect cystine in urine, the cyanide-nitroprusside reaction has been used as a clinical test [10]. Quantitative analysis methods for cystine have been reported using an amino acid analyzer, a HPLC method, thin-layer chromatography (TLC), and various modifications of colorimetric methods. Some methods are sensitive, but require special equipment (amino acid analyzer or HPLC), while other technique sacrifice sensitivity (TLC) [11]. After analyzing urinary cystine qualitatively and quantitatively, and repeating previously reported colorimetric methods [12–18], we were unable to produce reproducible results. We found that the causes of non-reproducible results were: (1) not enough time for reduction of cystine to cysteine, and (2) the presence of interference compounds (acetone, acetoacetate and creatinine) for color development [19]. The method reported here is based on the cyanide-nitroprusside reaction, modified for adequate reduction time of cystine to cysteine and for elimination of interferences. We tested other chemical compounds taken as daily supplements, i.e., ascorbic acid, iron (III) and zinc chlorides, to determine if these compounds interfered with our modified assay. Cystinuria patients are treated with captopril, thiola, or D-penicillamine to form a soluble mixed-disulfide complex with cysteine. We examined the effects of these drugs on cystine determination, and discovered that captopril produced color at 521 nm with increasing concentration, but both thiola and D-penicillamine did not develop color with increasing concentration.

2. Materials and methods Urine samples from 15 normal adults (10 males and five females, age 49.1

Y. Nakagawa, F.L. Coe / Clinica Chimica Acta 289 (1999) 57 – 68

59

610.2 years) and 12 cystine stone forming patients (five males and seven females, age 40.6 615.4 years), were collected at the Kidney Stone Laboratory of the University of Chicago. Each sample was collected for 24 h in a plastic bottle containing thymol as a preservative. Chemicals used were the highest grade available: ascorbic acid (Sigma), captopril (E.R. Squibb & Sons), creatinine (Sigma), L-(2)-cystine (Eastman Kodak), iron (III) chloride hexahydrate (Aldrich), N-(2-mercaptopropionyl)-glycine (Thiola, Sigma), D-Penicillamine (Sigma), human serum albumin (Sigma), sodium chloride (Fisher), sodium cyanide (Fisher), sodium nitroprusside (Sigma), monosodium phosphate (Mallincrodt), and zinc chloride (Fisher). Ascorbic acid (100 mg / ml), captopril (5.76 mg / ml), FeCl 3 .6H 2 O (10 mg / ml), penicillamine (3.60 mg / ml), thiola (5.14 mg / ml), and ZnCl 2 (10 mg / ml) were dissolved in double distilled water prior to use. All other reagents were prepared using double distilled water. As a control, Urine 1 and Urine 2 solutions from Ciba-Corning were used (now available through BioRad Laboratories). The control solutions were prepared by dissolving the dry powder into 25 ml of double distilled water at room temperature, as per the manufacturer’s instructions. Urine pH was measured by a Beckman Phi pH meter with a combination electrode. The pH meter was calibrated with reference solutions of pH 4 and 10 (Fisher Scientific), and standardized with pH 5 and 7 buffer solutions (Fisher Scientific). Creatinine concentration was determined using a Beckman CX-5 autoanalyzer. An aliquot of normal urine, 400 ml, was mixed with 100 ml of water and 1 ml of 0.01 M sodium phosphate buffer containing 0.15 mol / l NaCl, pH 7.3 (PBS). Following the addition of 300 ml of 10% NaCN aqueous solution (w / v), the samples were incubated for 20 min at room temperature. Color was developed by adding 100 ml of 20% sodium nitroprusside aqueous solution (w / v) and the color intensity was measured at 521 nm within 1 min using a Beckman spectrophotometer (model DU 640). (The final volume for the test was 1.9 ml). A blank was prepared with the equivalent creatinine concentration for each sample using a 1 mg / ml of creatinine aqueous stock solution. For cystinuria patients, 100 ml, were used instead of 400 ml of normal urine, along with 300 ml of water and the same amounts of the reagents as described above, to make the final volume of 1.9 ml. To determine the maximum saturation point of cystine, 100 mg of cystine crystal was added into 1 ml of urine specimen with thymol in a tightly sealed tube, and incubated at 388C for 48 h. A 100 ml aliquot of urine was used for cystine determination after centrifugation at 2500 rpm for 10 min (Beckman Spinchron Centrifuge). Five different concentrations, ranging from 0.4 mg to 3 mg, of captopril, penicillamine or thiola solution were tested the same manner for cystine quantitation. The extent of interference of ascorbic acid, iron and zinc was tested by adding these compounds into the assay solution. The stock solution of ascorbic acid was

60

Y. Nakagawa, F.L. Coe / Clinica Chimica Acta 289 (1999) 57 – 68

added into assay solutions to make a final concentration of 25 mg / ml ascorbic acid, or a final concentration of 1 mg / ml of FeCl 3 or ZnCl 2 into the each assay solution. This was followed by cystine determination as described above. Five cystinuric urine samples were analyzed by a Beckman amino acid analyzer (Model 121M). One ml of urine was adjusted to pH 10 using 100 ml of 1 mol / l NaOH and ammonia was removed in a vacuum dessicator over conc. H 2 SO 4 for 12 h. Since approximately 200 ml of water evaporated during removal of ammonia under vacuum, the sample was acidified to pH 2 with 50 ml of 4 mol / l HCl, the final volume was adjusted to 1 ml by adding water. The sample was then analyzed using an amino acid analyzer.

3. Results The absorbance at 521 nm was 0.60 after 10 min, 0.65 after 20 min, and 0.67 after 30 min incubation. Thus, the reduction of cystine to cysteine by NaCN required a 20 min incubation period at 308C in order to obtain the maximum color intensity (Fig. 1 and Reaction scheme). Fig. 2 shows the absorption spectra of controls containing different concentrations of cystine. The reaction product showed two absorption maxima, 400 and 521 nm. As previous reports have indicated, the absorbance at 521 nm exhibited the cysteine-nitroprusside complex. Fig. 3 shows the linear relationship of absorbance at 521 nm with increasing concentration of cystine from 5 mg / ml to 50 mg / ml per assay. Microalbuminuria in diabetic nephropathy is defined as 20 | 70 mg / min excretion of albumin in urine [20]. A normal value is quoted as 5.161.0

Fig. 1. Reduction time of cystine by NaCN and color intensity.

Y. Nakagawa, F.L. Coe / Clinica Chimica Acta 289 (1999) 57 – 68

61

Reaction scheme.

mg / min. Based on these values, the normal value is calculated as 7.3 mg / 24 h and the highest value of microalbuminuria is 100.8 mg / 24 h. The effect of albumin on cystine determination was examined using 100 mg of human serum albumin (Fig. 4). No effect was found at 521 nm; however, the interference of creatinine was significant (Fig. 5). The absorption maximum was 410 nm, but its absorbance at 521 nm was large. Also, this absorbance increased with time by reacting with sodium nitroprusside. The color intensity of cysteine starts decreasing after 2 min; however, the color intensity of creatinine keeps increasing with time. Thus, it is critical to limit the time to within 1 min after adding the chromogenic reagent. The interference of ascorbic acid, FeCl 3 and ZnCl 2 was tested at various concentrations of cystine. The results are summarized in Fig. 6. Ascorbic and FeCl 3 interfered with this assay, but ZnCl 2 at 2 mg / assay did not interfere. The effects of captopril, penicillamine or thiola were tested on this assay. As shown in Fig. 7, there was a linear relationship between color intensity and increasing concentration of captopril. However, penicillamine and thiola reached a maximum absorbance of 0.075 at 521 nm with 0.5 mg, and after that no increase was observed with increasing concentration. When 0.5 mg of either

Fig. 2. Absorbance of cystine-nitroprusside complex with varying cystine concentration.

62

Y. Nakagawa, F.L. Coe / Clinica Chimica Acta 289 (1999) 57 – 68

Fig. 3. Calibration of cystine concentration.

penicillamine or thiola was added to urine, the absorbance at 521 nm was increased by only 0.075 absorbance units. Table 1 shows the creatinine concentration in urine specimens from normal (111.64659.63 mg / dl) and cystinuric patients (73.18642.08 mg / dl). Cystine content, after correction for interference due to creatinine, were (39.96623.83 mg / ml) for normals and (293.436156.47 mg / ml) for patients. After the urine

Fig. 4. Effect of human serum albumin on cystine determination.

Y. Nakagawa, F.L. Coe / Clinica Chimica Acta 289 (1999) 57 – 68

63

Fig. 5. Effect of creatinine on cystine determination.

was saturated with cystine, the cystine concentration of the patients urine increased to (395.19693.85 mg / ml) compared to (270.30659.51 mg / ml) for normals as shown in Table 1. Two patient urine samples were over 500 mg / ml after saturation with cystine. Since the solubility of cystine is extremely sensitive to pH, cystine stone forming patient urine showed higher solubility of cystine

Fig. 6. Interferences of cystine determination by dietary supplements.

64

Y. Nakagawa, F.L. Coe / Clinica Chimica Acta 289 (1999) 57 – 68

Fig. 7. Color reactions of D-penicillamine, thiola, and captoril with NaCN-Na nitroprusside reagent.

due to the higher pH of their urine compared to normals. Four urine samples, 2 normals and 2 stone formers, were adjusted to pH | 6.0 with 1 mol / l HCl, and cystine concentration was determined after saturation with cystine for 48 h at 378C. All cystine concentrations were within normal solubility ranges ( | 260 mg / ml) (Table 2). Using commercially available controls, cystine content was measured after correction for the interference of creatinine. The results are summarized in Table 3. The cystine concentrations in both samples showed agreement with expected values.

4. Discussion Cystine was first detected in cystinuria by Brand et al. using 5% NaCN for 10 min reduction of cystine [10]. Several investigators have modified this qualitative method for quantitative analysis of cystine using NaCN [11–13,15] or KCN [16] as a reducing reagent and a reaction time of 3, 5 or 10 min. Our results showed non-reproducibility after these short reduction times, even though we strictly timed the reduction and color developing reactions. As shown in Fig. 1, complete reduction of cystine by NaCN takes at least 20 min, and when the reduction time was extended to 30 or 40 min, the color intensity developed by

Y. Nakagawa, F.L. Coe / Clinica Chimica Acta 289 (1999) 57 – 68

65

Table 1 Cystine concentration in urine and after saturated with cystine at 378C for 48 h Name

Sex

pH

Creatinine mg / dl

Cystine, mg / ml Original

Normal urine L. S. P. F. M. D. P. M. B. C. V. L. R. C. F. G. F. D. E. A. N. N. B. P. L.H. L. G. K. F.

M M M M M M M M M M F F F F F

6.35 6.755 6.687 6.186 6.056 5.813 6.107 6.568 7.051 5.506 5.886 5.704 5.845 5.716 6.913

123.55 113.50 109.19 200.96 134.57 85.41 82.74 118.29 47.02 57.06 128.54 67.22 123.55 260.29 22.76

Cystinuria T. S.a J. H.a M. P. C. S.b C. G. C. O. M. D.a L. S. S. H. M. A. M. H. P. W.

M M M M M F F F F F F F

7.556 7.556 6.583 8.028 7.782 7.666 7.532 8.521 6.901 7.330 7.802 7.524

135.91 135.90 158.62 73.10 29.90 34.73 67.07 36.50 48.21 110.60 44.81 87.50

43.55 30.89 33.3 98.58 81.33 42.11 31.38 21.89 16.17 22.54 32.25 33.57 58.95 43.14 9.72 504.29 (550.08)c 187.64 (193.92)c 300.73 679.41 209.98 211.65 (222.71)c 325.94 (299.52)c 188.06 102.60 276.33 248.19 286.30 (360.96)c

After Cys. Sat. 217.4 308.94 275.94 392.62 355.27 264.06 236.56 206.88 245.16 224.91 205.25 233.31 256.96 378.40 262.96

553.15 353.64 351.82 620.89 341.82 395.69 400.11 346.25 327.71 318.32 365.39 367.49

a

Indicates treating with thiola. Indicates treating with captopril. c Values shown in parenthesis are cystine concentration in the original urine by amino acid analysis. b

nitroprusside was constant. During the course of this study, other reducing methods and chromogenic reactions for quantifying urinary cystine concentration were evaluated (data not shown). Roston [17], and Schneider et al. [18] developed color using norepinepherine to form a ‘‘noradrenochrome’’. Rosenthal and Yaseen proposed reduction of cystine with sodium borohydride [12], and Thuy and Nyhan [14] reported reduction of cystine by nascent hydrogen gas formed by mixing zinc in 6 mol / l HCl followed by sodium nitroprusside

Y. Nakagawa, F.L. Coe / Clinica Chimica Acta 289 (1999) 57 – 68

66

Table 2 The effect of pH on solubility of cystine in urine Name

Sex

pH Initial

Cystine concentration (ug/ml) of pH adjusted urine Adjusted Before sat. w/ cystine

Normal M. O. L. G.

6.892 6.282

5.972 5.972

27.20 31.22

240.13 261.8

Cystine Stone Forming Patients M. D.a F 7.532 C. S.b M 8.028

6.536 5.958

325.94 373.07

276.46 318.21

a b

M F

After sat. w/ cystine

Indicates treating with thiola. Indicates treating with captopril.

reaction. None of these methods had satisfactory performance. (The recovery of cystine was 75% of the theoretical values). The cause of unreliable results was thought to be due to other components in urine. We checked possible interferents, for example, albumin, since human serum albumin contains 35 residues of cysteine. The reduction of serum albumin by NaCN did not take place as shown in Fig. 4, thus we assumed that all proteins in urine would behave the same as albumin. It was discovered that creatinine caused serious interference with cystine determination. After treating creatinine in the same manner as cystine, a brown color developed immediately. The absorption maximum of the creatinine-nitroprusside complex appeared at 407 nm, and a broad shoulder was observed around 500 nm (Fig. 5), thus the maximum absorbance of cysteine-nitroprusside complex at 521 nm was influenced by this shoulder peak. The color intensity due to creatinine increased with time; however, the cystine complex color lasted for 2 min, then gradually disappeared. Thus it is important to measure the color intensity of the cysteine complex and that of creatinine complex at the same time interval after addition of the chromogenic reagent. In this report we set the interval at 1 min for subtraction of the creatinine background. This determination method was validated by using pre-assayed commercial controls producing cystine values Table 3 Assigned and / or observed cystine and creatinine concentration in control samples a Analyte

Value

Cystine, mg / l

Creatinine, mg / l a

Urine Control 1

Urine Control 2

Assigned mean (2 SD) Observed mean (n 5 6)

27 (16 | 32) 2867

15 (9 | 21) 1564

Assigned mean (2 SD) Observed mean (n 5 6)

99 (89 | 109) 9562

244 (220 | 268) 22665

Commercial controls from Ciba-Corning Corporation.

Y. Nakagawa, F.L. Coe / Clinica Chimica Acta 289 (1999) 57 – 68

67

that agreed with the package inserts (Table 3). Uric acid and ammonia did not interfere with this reaction as observed previously [10]. Cystinuric patients are treated medically with captopril, n-(2-Mercaptopropionyl)glycine (thiola), or penicillamine to convert less soluble cystine to the more soluble cysteine-drug disulfide complex. As shown in Fig. 7, captopril gave linear color development with increasing concentration; however, thiola and penicillamine did not develop color with increasing concentration. One of the cystine stone-forming patients (C.S.) who was treated with captopril showed high cystine content in his urine (679.41 mg / ml) at pH 8.028 (Table 1), after adjusting pH to | 6.0, cystine content was still 373.07 mg / ml, and after saturating with cystine, the cystine value was 318.21 mg / ml (Table 2). These high values indicate clearly captoril interfered with the reaction and gave higher values. In contrast, four patients were treated with thiola, and cystine values were not extremely different with patients who were not treated with medications. Cystine value decreased in one of these patients (M. D.) from 325.94 mg / ml to 201.1 mg / ml after the pH was changed from 7.5 to 6.5. The cystine value of this urine (at pH 6.5) after saturated with cystine was 276.46 mg / ml which was an expected value at this pH as reported previously (Table 2). This indicated that thiola-cysteine disulfide and thiola disulfide, or penicillamine complexes were not reduced by NaCN. Previously, Lotz et al. reported that penicillamine-cysteine disulfide and penicillamine disulfide did not develop color by the NaCN-nitroprusside reagent [21]. The amounts of cystine in five cystinuric patients’ urine were verified using an amino acid analyzer. The values of four out of five agreed within 10% of the value obtained by this colorimetric method, and only one case yielded 26% less than the value obtained by the amino acid analyzer. The cause of the discrepancy of cystine values between NaCN-nitroprusside colorimetric method and amino acid analysis is currently under investigation. Ascorbic acid and iron (III) chloride, commonly used as dietary supplements interfered with the assay. Thus, these daily supplements must be eliminated prior to cystine analysis. Our results suggest that the previously reported reduction time of cystine using NaCN is not sufficient, and that it requires at least 20 instead of 10 min for the reaction to go to completion. In general, this colorimetric method has advantages over other methods in time, cost and simplicity of technique.

Acknowledgements This research was supported by the National Institute of Diabetes and Digestive and Kidney Diseases through the O’Brien Kidney Research Center Grant P50-DK-47631. The authors wish to thank Dr. G. A. Reddy for conducting the amino acid analyses.

68

Y. Nakagawa, F.L. Coe / Clinica Chimica Acta 289 (1999) 57 – 68

References [1] Miller DS. Cystinuria. Endocrinol Metabol Clin N Am 1990;19:889–907. [2] Pras E, Arber N, Aksentijevich I et al. Localization of a gene causing cystinuria to chromosome 2p. Nature Genet 1994;6:415–9. [3] Calonge MJ, Gasparini P, Chillaron J et al. Cystinuria caused by mutations in rBAT, a gene involved in the transport of cystine. Nature Genet 1994;6:420–5. [4] Miyamoto K, Katai Tatsumi S et al. Mutations of the basic amino acid transporter gene associated with cystinuria. Biochem J 1995;310:951–5. [5] Endsley JK, Phillips III JA, Hruska KA et al. Genomic organization of a human cystine transporter gene SLC3A1) an identification of novel mutations causing cystinuria. Kidney Int 1997;51:1893–9. [6] Horsford J, Saadi I, Raelson J, Goodyer PR, Rozen R. Molecular genetics of cystinuria in French Canadians: identification of four novel mutations in Type I patients. Kidney Int 1996;49:1401–6. [7] Gitomer WL, Pak CYS. Recent advances in the biochemical and molecular biological basis of cystinuria. J Urol 1996;156:1907–12. [8] Saadi I, Chen XZ, Hediger M et al. Molecular genetics of cystinuria mutation analysis of Slc3a1 and evidence of another gene in the Type I (silent) phenotype. Kidney Int 1998;:48– 55. [9] Goodyer P, Saadi I, Ong P, Elkas G, Rozen R. Cystinuria subtype and the risk of nephrolithiasis. Kidney Int 1998;54:56–61. [10] Brand E, Harris MM, Biloon S. Cystinuria. The excretion of a cystine complex which decomposes in the urine with the liberation of free cystine. J Biol Chem 1930;86:315–31. [11] Giugliani R, Ferrari I, Greene LJ. An evaluation of four methods for detection of heterozygous cystinuria. Clin Chim Acta 1987;164:227–33. [12] Rosenthal AF, Yaseen A. Improved qualitative screening test for cystinuria and homocystinuria. Clin Chim Acta 1969;26:363–4. [13] Georges XJ, Politzer WM. pH dependency of the cyanide-nitroprusside colour reaction of cystine. Clin Chim Acta 1970;30:737–40. [14] Thuy Le P, Nyhan WL. A screening method for cystine and homocystine in urine. Clin Chim Acta 1992;211:175–9. [15] Wu JT, Wilson LW, Christensen S. Conversion of a qualitative screening test to a quantititative measurement of urinary cystine and homocystine. Ann Clin Lab Sci 1992;22:18–29. [16] Matsuura T, Szafarz D. Micro-method for the estimation of cystine in protein hydrolysates. Experientia 1965;21:737–8. [17] Roston S. Determination of cysteine. Anal Biochem 1963;6:486–90. [18] Schnieder JA, Bradley KH, Seenmiller JE. Colorimetric assay of cystine using noradrenochrome. Anal Biochem 1968;23:129–31. [19] Greenstein JP, Winitz M. Cystine and cysteine. In: Chemistry of the Amino Acids, New York: John Wiley & Sons, 1961, pp. 1879–928. [20] Mogensen CE. Microalbuminuria as a predictor of clinical diabetic nephropathy. Kidney Int 1987;31:673–89. [21] Lotz M, Potts JT, Bartter FC. Rapid simple method for determining effectiveness of D-penicillamine therapy in cystinuria. Brit Med J 1965;5460:521.